Culture-independent identification of gut bacteria correlated with the onset of diabetes in a rat model

Abstract

Bacteria associated with the onset of type 1 diabetes in a rat model system were identified. In two experiments, stool samples were collected at three time points after birth from bio-breeding diabetes-prone (BB-DP) and bio-breeding diabetes-resistant (BB-DR) rats. DNA was isolated from these samples and the 16S rRNA gene was amplified using universal primer sets. In the first experiment, bands specific to BB-DP and BB-DR genotypes were identified by automated ribosomal intergenic spacer analysis at the time of diabetes onset in BB-DP. Lactobacillus and Bacteroides strains were identified in the BB-DR- and BB-DP-specific bands, respectively. Sanger sequencing showed that the BB-DP and BB-DR bacterial communities differed significantly but too few reads were available to identify significant differences at the genus or species levels. A second experiment confirmed these results using higher throughput pyrosequencing and quantitative PCR of 16S rRNA with more rats per genotype. An average of 4541 and 3381 16S rRNA bacterial reads were obtained from each of the 10 BB-DR and 10 BB-DP samples collected at time of diabetes onset. Nine genera were more abundant in BB-DP whereas another nine genera were more abundant in BB-DR. Thirteen and eleven species were more abundant in BB-DP and BB-DR, respectively. An average of 23% and 10% of all reads could be classified at the genus and species levels, respectively. Quantitative PCR verified the higher abundance of Lactobacillus and Bifidobacterium in the BB-DR samples. Whether these changes are caused by diabetes or are involved in the development of the disease is unknown.

Figures

Figure 1

Gel-like image generated by the bioanalyzer. The first and last columns are the reference DNA 7500 ladder. Base pair sizes are indicated adjacent to the ladder. Samples 1–3 represent the automated ribosomal intergenic spacer analysis (ARISA) profiles for the intestinal tract of bio-breeding diabetes-prone (BB-DP) rats and samples 4–6 represent bio-breeding diabetes-resistant (BB-DR) rats' stool samples at 60 days of age. The lowermost (50 bp) and the uppermost (10 380 bp) bands represent the markers used to align the ladder data with data from the sample wells. The boxes represent dominant bands unique to both groups and were extracted from the gel for further sequencing. The stool samples used in the ARISA come from experiment 1.

Figure 2

Principal coordinates analysis (PCA) depicting the qualitative (presence/absence) and quantitative (presence/absence and abundance) of the bacterial communities for the 10 stool samples each from the bio-breeding diabetes-resistant and bio-breeding diabetes-prone rats. This analysis is based on the community structures derived from Sanger sequencing of experiment 1 (A) and pyrosequencing of experiment 2 (B).

Figure 3

Shannon–Weaver and richness diversity indices calculated for the three time points after birth when stool was collected. Circles and squares represent the BB-DP and BB-DR samples, respectively. Open symbols represent the richness index (d) whereas closed symbols depict the Shannon–Weaver (H′) indices. Indices were calculated using automated ribosomal intergenic spacer analysis (ARISA) data from the experiment 2 samples.

Figure 4

Log of the number of Lactobacillus and Bifidobacterium cells per 5 ng of DNA from bio-breeding diabetes-resistant (BB-DR) and bio-breeding diabetes-prone (BB-DP) stool samples. (a) Experiment 1 (three stool samples per genotype). (b) Experiment 2 (10 stool samples per genotype). The standard error about the mean is depicted in the error bar about the data columns.

Figure 5

Family-level phylogenetic classification of those OTUs that could not be classified at the genus or species levels. Red branches depict 16S rRNA sequences from bio-breeding diabetes-prone (BB-DP) rats and green branches depict sequences from bio-breeding diabetes-resistant (BB-DR) rats. Branches in black depict known sequences from bacterial isolates. A list of bacterial isolates that were aligned with the sequences obtained from this study can be seen in the supplementary material. Sequences were aligned by using NAST (DeSantis et al., 2006), the aligned sequences and its respective nearest isolates were uploaded in MEGA 4 (Tamura et al., 2007) for conduction of the phylogenetic analysis. The evolutionary history was inferred using the neighbor-joining method and the evolutionary distances were computed using the maximum composite likelihood method. All positions containing gaps and missing data were eliminated from the data set (complete deletion option). Striking taxonomic trends were observed with the Clostridiaceae and Ruminococcaceae more prevalent in BB-DP whereas the Lachnospiraceae, Porphyromonadaceae and Prevotellaceae were more common in BB-DR.